hard drive related reiserfs crash

[dendenners]

Hi,
I have a problem with a machine running gentoo. Every so often (roughly every 24 hours or so) the machine stops responding to commands (for example when I run programs I often get errors like 'I/O error' or errors relating to the bus) and in the kernel logs I get messages like
end_request: I/O error, cmd 0 dev 03:08(hda) sector 183624
zam_7001: io error in reiserfs_find_entry
The only way I can restart the machine when this happens is to hard restart it.
A number of different sectors come up in the error message at different times. I ran fsck on all the partitions I have, but fsck reckons everything is OK. I am running oracle 9 on the machine - this may have nothing to do with the issue, but I can't recall the problem occuring when oracle was not running on the box. What could be the problem here? Is the hardware dodgy (as I suspect) or are oracle's onerous memory requirements leading to unforseen errors elsewhere? I'd be more than grateful for any pointers
Regards
Denis

[Janne Pikkarainen]

Sounds like a soon-to-be-dead HD. If I were you, I would replace the HD as soon as possible.
_________________
Yes, I'm the man. Now it's your turn to decide if I meant "Yes, I'm the male." or "Yes, I am the Unix Manual Page.".

[perm]

You could try to run one of the S.M.A.R.T. tools (emerge ide-smart) and see if any of its tests fails. If it does, you should already have made a backup and ordered a new drive. Alternatively, tar down everything to somewhere else, and reformat the partition in question to e.g. ext3. This doesn't really help if you've got a bad drive, but it might if you've run into some weird reiserfs-related problem.

On second thought, you don't have some cron-job giving the the drive a workout once a day, so that it gets too hot?

[dendenners]

OK guys, I've got more info on this. The following data is printed to the console just as the box abends: